To date, the majority of recommender systems (RSs) work on a single domain, such as exclusively for movies, books, etc. However, human preferences may span across multiple domains. Hence, consumption behaviors on related items from different domains can be useful to inform RS to make recommendations. This paper reports our efforts on uncovering the association between user preferences on related items across domains. In addition, we have also tested collaborative filtering technique on our cross-domain dataset for which results are reported here.